From low-quality computer-written books flooding the market to potential copyright violations, publishing is the latest industry to feel the threat from rapid developments in artificial intelligence. From writing and editing to distributing, the technology is transforming every aspect of the business. Some see it as a threat, while others embrace its possibilities.
Generative AI — which can produce input-based text — is the most problematic for many publishers and writers. It’s already used to write novels but can also be adapted to other genres, including poetry. For example, a poet known as David Jhave Johnston recently released a limited edition box set of 12 poems that were generated by an AI program and edited by him.
The new tech has sparked fears that it could undermine traditional publishing and destroy its labor-intensive aspects, such as editing and proofreading. A writer’s group executive director says if the technology becomes advanced enough, it could turn book writing from “a craft into a commodity,” leaving writers out of work. “People will start using this tool to create the whole thing from beginning to end, and we’re going to be in trouble,” she told The Verge. Several tech entrepreneurs have pinged the book world’s radar and raised its hackles by hawking AI “solutions” that will automate the writing process. But most of those treats writers like an afterthought, such as Sudowrite’s Story Engine (dubbed the “AI novel-writing tool everyone hates”).
Other technologies, such as neural networks and machine learning algorithms, are a less direct threat because they aren’t meant to replace human jobs but to enhance them. In the publishing world, for instance, a neural network can analyze large amounts of data to find trends in sales figures and reader demographics. It can then help publishers and retailers develop targeted marketing strategies.
But despite the potential of these new technologies, there’s a sense among many book industry players that it is too soon to know how profoundly they will affect their businesses. They worry that the data quality fed to generative AI needs to be higher to ensure accuracy and usefulness.
They also wonder how new content created by generative AI will be vetted and whether it will comply with existing intellectual property law. It may be a while before these questions are answered, but it is clear that a new era of book publishing has begun.
AI is already reshaping how the book industry works, from composing and writing to publishing, selling, and promoting. It’s being incorporated into everything from editing manuscripts to analyzing consumer data and making recommendations to readers.
However, many barriers must be overcome before generative AI can entirely disrupt the publishing industry. The most important one is probably the quality of the data being fed to the AI — without it, the technology won’t be as accurate or adequate. The data includes not just sales and reading statistics but a variety of other details about each book, such as its title, cover art, and contents.